Live data from Hacker News

Slick – Carousel

kenwheeler.github.io

31–40 of 119 posts

Re: Slick – Carousel

#32

I don't wish to engage in standard HN negativity, but I'd like to suggest that you should probably don't want to use a carousel at all most of the time. From the user's perspective, a carousel is just a random image. You'll get more conversions if you choose the best image instead of a random one. http://www.bayesianwitch.com/blog/2014/carouselsarebadforcon...

I honestly can't think of an instance where a carousel is the best UI element to use, yet everyone requests them. Why?

Exactly. The carousels on The Verge and Polygon are just awful. The only good implementation of a carousel on mobile (in my experience) is the one on GSMArena, which has large prev/next buttons.

Re: Slick – Carousel

#33

I don't wish to engage in standard HN negativity, but I'd like to suggest that you should probably don't want to use a carousel at all most of the time. From the user's perspective, a carousel is just a random image. You'll get more conversions if you choose the best image instead of a random one. http://www.bayesianwitch.com/blog/2014/carouselsarebadforcon...

The maths looks wrong; the distribution of clicks is only relevant if the total number of clicks stays the same.

Essentially a carousel gives you a "second-bite" at your customer, i.e. the second image is a chance to convert those customers who haven't already been converted by the first image.

Imagine you have two images one which converts at 40% and one which converts at 20% - but each image converts completely different groups; by carouseling the images you could end up with 40%+20% conversion.

In practice you won't because of overlaps, drop-off from people who never see the second image, etc. but it's definitely possible to exceed the maximum conversion of an individual image.

(obviously you should A/B test for your circumstances)

Re: Slick – Carousel

#34

I don't wish to engage in standard HN negativity, but I'd like to suggest that you should probably don't want to use a carousel at all most of the time. From the user's perspective, a carousel is just a random image. You'll get more conversions if you choose the best image instead of a random one. http://www.bayesianwitch.com/blog/2014/carouselsarebadforcon...

I honestly can't think of an instance where a carousel is the best UI element to use, yet everyone requests them. Why?

They're an absolute boon to Web designers where different stakeholders in the client organisation can't make up their bloody minds. There comes a snapping point where a carousel is suggested to avoid insanity.

Re: Slick – Carousel

#36

I don't wish to engage in standard HN negativity, but I'd like to suggest that you should probably don't want to use a carousel at all most of the time. From the user's perspective, a carousel is just a random image. You'll get more conversions if you choose the best image instead of a random one. http://www.bayesianwitch.com/blog/2014/carouselsarebadforcon...

Highly relevant: http://shouldiuseacarousel.com/

Re: Slick – Carousel

#37
It looks to me like pinch/zoom doesn't work. Not a big deal, but I wouldn't call it perfect until it's flawless on a tablet and phone.

Re: Slick – Carousel

#38
Testing it out as a user, the slide transitions didn't feel as responsive to swipe velocity as I expected. In particular, if I swipe even a small distance but quickly, I expect to trigger the transition. I just tested it with iOS and Android, and they both seem to do it at least for left-to-right transitions of home screen pages, and Android does it for the top menu.

From looking at [the source](https://github.com/kenwheeler/slick/blob/master/slick/slick....), it doesn't seem check for velocity; it just applies a distance threshold. Have you considered having the threshold respond to swipe speed so that faster swipes over shorter distances trigger transitions?

Re: Slick – Carousel

#40

It's using jQuery, so no. It'll not be the last one I'll ever need.

Have you ever used a carousel on a site where you didn't use jQuery? Would you prefer the size of this library triples? Do you think any one man has the time to keep up with the kind of event normalization and browser shimming that jQuery provides?
Post reply on HN